SPE-172633-MS

Application of Fast Reservoir Simulation Methods to Optimize Productionby Reallocation of Water Injection Rates in an Omani Field

A. Al Saidi, P. Pourafshary, and M. Al Wadhahi, Department of Petroleum and Chemical Engineering, SultanQaboos University

Copyright 2015, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Middle East Oil & Gas Show and Conference held in Manama, Bahrain, 811 March 2015.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contentsof the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflectany position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the writtenconsent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations maynot be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract

There are wide and crucial applications of the reservoir simulation to develop, manage and monitorreservoirs. However, the traditional simulation methods have different limitations such as noticeablenecessary time to prepare and model data and uncertainties in the available data. Therefore, newapproaches are developed to offer rapid reservoir simulation with lower and more reliable data to manageand optimize production from hydrocarbon reservoirs and provide us a quick estimation of the perfor-mance of the reservoirs. One of these approaches is the Capacitance-Resistive Model (CRM).

CRM is a quantitative technique based on material balance that uses only injection/production rates andwell coordinates to identify and quantify reservoir model parameters, which are well connectivity and timeconstant values. In CRM, the reservoir is assumed as a system where the injection rates are the inputsignals and the production rates are the output signals. The injection performance over a period of timecan be identified by estimating fractions of injected fluid to each producer and the time taken for aproducer to sense the injections. With CRM, it is possible to optimize a waterflooding operation in a fieldby reallocating the water injection rates to optimize the oil production. This tool is preferred due to itssimplicity, the short computation time and the use of available field data.

In this study, we developed a new approach to use the Capacitance Resistive method to optimize thewaterflooding process in an Omani oil field by a ranking approach for the injectors. By reallocation of theavailable water to injectors, the future oil production was optimized for the next three years. This methodshowed that the optimized scenario leads to 29.90% increase in future oil production compared to thecurrent injection rates profile. This was achieved by injecting 59% of available water for most effectivewells and the remaining amount for other wells.

IntroductionReservoir simulation is an essential technique to model, develop and manage reservoirs. It is used topredict future production behavior of an oil or gas reservoir under different scenarios of production andinjection. Hence, reservoir simulation is an important tool for reservoir management to enhance theproduction of oil and gas.

The traditional approach of reservoir simulation is based on modeling the continuity equation andDarcys law for flow through porous media. These methods have major limitations in their applications,for example, preparation of a large amount of reservoir data, modeling and analyzing results are timeconsuming by most of such methods especially for the large fields. It may take months to set up andcomplete one simulation model. Therefore, in many situations running full-scale numerical simulationsdoes not meet the economical requirement and the time limit restrictions of the project. Another problemfor the traditional methods is uncertain data that are used in the simulation, which leads to errors in thecalculations.

Capacitance Resistive Model (CRM) approach is a promising rapid evaluator of reservoir performance,which has been recently used to manage and optimize production from hydrocarbon reservoirs. It is aquantitative technique based on material balance that uses only injection/production rates and wellcoordinates to identify and quantify interwell connectivity values in waterflooding. CRM does not requiregeological information and considers the reservoir as a system that has injection rates as the input andproduction rates as the output (Figure 1). This tool is preferred due to its simplicity, the short computationtime, and the use of available field injection/production data. However, it has disadvantages regarding thesensitivity of reservoir events and data errors.

The idea of applying a linear multivariate regression technique to predict the production of a well basedon injection rates was constructed by Albertoni and Lake (2003) who used a simple model to deduct theinterwell connectivity in a reservoir using only injection and production data. Gentil et al. (2005)demonstrated the physical meaning of the production as the ratio of transmissibility in a producer-injectorchannel over the total transmissibility of the entire area surrounding an injector. In 2006, Yousef et al.developed a more complicated model that relies on the material balance that consists of capacitance(compressibility) and resistive (transmissibility) effects. They named their model the Capacitance Resis-tive Model. For each injector-producer pair, they specified two coefficients: the weight identifies theconnectivity and the time constant identifies the degree of fluid storage around each producer.

Figure 1Schematic representation of the impact of an injection rate signal on total production response for an arbitrary reservoircontrol volume in the CRM (Sayarpour, 2008)

2 SPE-172633-MS

If we assume that there is a volumetric balance between each pair of injector/producer in the reservoiras shown in Figure 1, the following differential equation shows the relation for the flow rate from injectori to producer j (qij) as a function of CRM parameters which are well connectivity between injector i andproducer j (fij) and time constant value (ij) (Sayarpour et al. 2009).

(1)

where time constant is defined as , i 1,2,. . ., Ni and j 1,2,. . ., Np

The production rate for producer j will be the summation of all qij values, calculated by solvingequation (1), which is equal to:

(2)

Where and represent injection rate of injectior i and changes in bottomhole pressure (BHP)

of producer j during time interval tk-1 to tk, respectively.The developed model can be used to maximize oil production by reallocating available amount of water

to the injectors which is an approach for waterflooding optimization. Sayarpour et al. (2009) applied theCRM approach for real fields to validate the performance of CRM for waterflooding simulation. Weberet al. (2009) used this model to optimize waterflooding in large scale fields. They suggested a practicaltechnique to pre-screen the data to remove inactive wells and outliers which can strongly influence theresulting model fits.

Nguyen et al. (2011) developed an integrated capacitance resistance model (ICRM) that fittedcumulative production against rates, BHP, and cumulative water injected. Salazar (2012) presented a casestudy of CRM application combined with decline-curve analysis to successfully predict the behavior ofa mature reservoir under gas injection. Mamghaderi and Pourafshary (2013) extended the available CRMapproach to estimate and optimize waterflooding performance in layered reservoirs by coupling CRM andPLT data. In this paper, we will present the application of CRM in order to optimize waterflooding in anOmani oil field by reallocating the available water to injectors.

Omani field case studyThis Omani Oil Field is located in South Oman salt basin. It comprises a NE-SW trending anticline about14 km long and 8 km wide. Oil in this field has moderate to low API gravity (22 deg API) with viscosityof 90 cp. The original reservoir pressure was 9300 Kpa. Porosities range from 25% to 30%. Thepermeability of the reservoirs varies widely from about 100 md to over 5 Darcies. Figure 2 represents thewells location with wells co-ordinates. It shows that the field consists of 7 injectors and 26 producers.

SPE-172633-MS 3

The water injection data, liquid production rate, oil production rate, and water cut from all wells werecollected to develop the CRM model for the field. Outliers for the production/injection history wereremoved from the raw data to enhance the accuracy of the developed CRM model by reducing the unusualfluctuations. These outliers appeared in the history due to the operational reasons and testing operationson well. Moreover, data are prepared by applying a smoothing procedure. This reservoir is supported bytwo aquifers at the South East and North side of the field. Measurements in the field showed that 75% ofthe production is due to the support from injectors and the remaining is because of the water influx fromaquifers. We modeled the effect of aquifers by adding virtual injection signals at the location of aquifers.

CRM parameters for each pair of injector/producer as well as the aquifer parameters are calculated byminimizing the error between the real data and estimated results. Table 1 represents CRM parameters foreach well in the field. The higher value for f shows greater connection between the injector and theproducer. In addition, a large time constant indicates weak or no connection between the wells in the field.Some producers are affected only by the Northern aquifer such as P23 and P25. Also, there are producersthat affected by the South East aquifer like P3 and P13. Moreover, P9 and P20 are examples of theproducers which affected by both aquifers while P2, P6 and P21 are not affected by the aquifers.

Figure 2A schematic of the wells in the Omani Field

4 SPE-172633-MS

Figure 3 shows real and simulated production history for two wells as an example. As shown, there isa good agreement between the model and the real history. The average error for the developed model forthe whole reservoir history simulation was around 8.7%.

Table 1CRM coefficients for the Omani Field

SPE-172633-MS 5

CRM method provides the total liquid production from the wells. To optimize the oil production, amodel is needed to calculate the oil fraction for each well. In this work, an approach based on water cuthistory is used to simulate the oil production from liquid production in the future. Water cut history iscollected for each producer and used to develop a model and predict the future water cut. The data for therecent history is used to minimize the effect of operations on the water cut alteration. The future oilfraction is calculated from the predicted water cut fraction.

Each injector has different effects on the overall production from the field. Some of the injectors arevery active and lead to increase in production noticeably, but some of them are drilled at unsuitablelocations. Hence, most of the injected water does not reach and affect the producers. For example,presence of a fracture or high permeable media between an injector and a producer may lead to earlybreakthrough of the water and reduce the performance of waterflooding. In this research, a new approachis developed to study the effect of each injector on the producers to rank the performance of the injectors.From the relationship between the total oil production for all producers and the injected amount to aspecific injector, the curves are drawn to rank the injectors based on their effect on the oil production asshown in Figure 4.

Figure 3Matching in producers P10 and P15

6 SPE-172633-MS

Figure 4 shows that we have different types of injectors in the field based on their performance on theoil production enhancement. Hence, we divided the injectors into three categories as shown in Table 2.

The objective function of optimization is to optimize the future oil production for the next three yearswith two constrains. First, 1000 m3 is the maximum injected amount per injector. Second, 3372 m3 is theamount of available water for injection. Both constrains are based on the historical data. By ouroptimization algorithm, more water is allocated to good and normal wells and less water to the weak ones.By this approach, oil production from a field can be increased by reallocation of the available water basedon the connectivity values between the well pairs. Since injectors I1 and I6 are good injectors, more waterwill be injected to them and the rest will be distributed among the other injectors. Different scenarios aretested to reach an optimized scenario for the injection profile and an optimum oil recovery. Injecting 59%of available water to good injectors and 41% to normal injectors is the optimized case for the given fieldsince it has 29.90% increase in future oil production more than base case which refers to keep injectingsame as the current rate (Figure 5).

Figure 4Ranking results

Table 2Categories of injectors based on the ranking approach

SPE-172633-MS 7

ConclusionCapacitance Resistive Model (CRM) was used as a rapid reservoir simulation tool to manage and optimizethe production from reservoirs. It used only injection/production rates and well coordinates to identify andquantify interwell connectivity parameters during waterflooding in a reservoir. By estimating fractions ofinjected fluid and the time taken to reach a producer, CRM parameters were calculated and then used topredict and optimize oil production by reallocating the available water to the injectors.

In this study, CRM was applied to an Omani oil field to forecast and optimize the future productionby reallocating the injection profile to the injectors. Using historical data for the given field, CRMparameters were determined by minimizing the error between real data and CRM results. The resultsshowed that the average error for matching was around 8.7%. The results of ranking approach showed thatI1 and I6 are good injectors and by injecting more water to them in the future, these wells can support theproducers for more production. The best approach of optimization (59% of available water is for goodinjectors and 41% for normal injectors) leads to 29.90% increase in future oil production more than usingthe base case.

AcknowledgementWe would like to express our sincere gratitude to Petroleum Development Oman (PDO) for funding thisproject as a Msc thesis. Also we would like to thank Mr. Biswajit Choudhuri for his crucial role ofproviding the required data to complete the thesis.

ReferencesAlbertoni, A. and Lake, L. W., Inferring Connectivity Only From Well-Rate Fluctuations in

Waterfloods, SPE Reservoir Evaluation and Engineering, 2003, 6, 616.Gentil, P. H., The Use of Multilinear Regression Models in Patterned Waterfloods: Physical Meaning

of the Regression Coefficients, M.S. Thesis, The University of Texas at Austin, 2005.Mamghaderi, A., & Pourafshary, P., Water flooding performance prediction in layered reservoirs

using improved capacitance-resistive model, Journal of Petroleum Science and Engineering,2013, 108, 107117.

Figure 5Best optimization scenario and cumulative oil production rate using optimized case

8 SPE-172633-MS

Nguyen, A. P., Kim, J. S., Lake, L. W., Edgar, T. F., & Haynes, B. Integrated Capacitance ResistiveModel for Reservoir Characterization in Primary and Secondary Recovery, In SPE AnnualTechnical Conference and Exhibition. USA, January, 2011.

Salazar, M., and Gonzalez, H. Combining Decline-Curve Analysis and Capacitance-ResistanceModels to Understand and Predict the Behavior of a Mature Naturally Fractured CarbonateReservoir under Gas Injection. In SPE Latin American and Caribbean Petroleum EngineeringConference. Mexico, April, 2012.

Sayarpour, M., Kabir, C. S., & Lake, L., Field applications of capacitance-resistance models inwaterfloods, SPE reservoir evaluation & engineering, 2009, 12(6), 853864.

Weber, D., Edgar, F. T, Lake, L. W., Lasdon, L. S., Sawas K., Improvements of CapacitanceResistive Modeling and Optimization of Large Scale Reservoirs, SPE Western Regional Meeting,USA, March, 2009.

Yousef, A.A., Gentil, P.H., Jensen, J.L. and Lake, L.W., A Capacitance Model to Infer InterwellConnectivity from Production and Injection Rate Fluctuations, SPE Reservoir Evaluation &Engineering, 2006, 9(5), 630646.

SPE-172633-MS 9

Application of Fast Reservoir Simulation Methods to Optimize Production by Reallocation of Water ...IntroductionOmani field case studyConclusion

AcknowledgementReferences